Motor Rehabilitation Based on Brain Machine Interface and Microsoft Kinect

Abstract

This paper presents a brain-machine interface (BMI)-based motor rehabilitation system for detecting the correct onset-time of motor intention by Microsoft KinectTM. BMI-based motor rehabilitation without time delay is a key challenge due to muscle spasticity of patient for detecting the motor intention in a real-time rehabilitation system. To circumvent the time delay problem, we employed Kinect on a BMI rehabilitation system which calibrates the onset-time of brain signal and assesses the behavioral pattern of motor activity. In this paper, we demonstrate that the onset-time calibration of brain response by the Kinect-BMI rehabilitation system can minimize time delay and estimate motor intention before the actual movement of the user. In addition, visual interaction through Kinect would increase engagement for a given task and could positively encourage motor recovery for patients.